chapter - 5 the input subsystem - ajay ardeshana · fundamentals of gis chap. 5 : the input...

Fundamentals of GIS Chap. 5 : The Input Subsystem

Prepared By: Ajay A. Ardeshana, Lecturer of MCA, GARDI VIDYAPITH, Rajkot.Email: [email protected] Mobile: +91-95588 20298

Page : 1

Primary Data :-

Primary Data are those that are developed by the individual or

organization that intend to use them or, in some cases, are contracted out

with specific guideline how they will be prepared.

Primary Data are generally going to experience a higher level of quality

control then secondary data. They may required both intensive collection

(Field Collection) and conversion (Digitizing) than secondary data.

As a result they are often better quality for the specific application and

more costly to produce.

There are many forms of primary data ranging :

o Field Observation

o Biological Specimen Sampling

o Human Interviews and Surveys

o Arial Photograph and Satellite Remote Sensing

o Field Mapping

o GPS Surveys

o Police Calls and Accident Reports etc…

It would require volume to summarize all the guidelines and criteria for

all possible primary data collection methods.

Input Device :-

1. Digitizer :-

For inputting the spatial data manually, the use of the digitizer is

standard.

The digitizer is somewhat more advance and much more accurate version

of an input device used in all modern PCs.

Inside the Mouse of Digitizer, the Sensors that are responds to the motion

of rubber ball encase in the frame of device.

All the locational information is contained within the mouse itself.

The electronically active grid inside the digitizing table will record the

degree of movement for the digitizer.

The device like mouse called puck is connected to the table and move

along the table and it moved to the different location on the map that is

attached to the table.

He digitizer puck contain the crosshair deice encased in glass or clear

plastic, that allows the operator to place the puck exactly over individual

map element.



Page : 2

The puck has a buttons that indicate the starting and ending of lines or

polygons or explicitly define left and right polygon and so on.

Modern digitizer can provide resolution of 0.001 inch, for an area of 42

inches by 60 inches.

Some of the useful factors are there for selecting a digitizer.

o Stability :-

Deal with the tendency of the exact reading of the digitizer to change as

the machine warms up. The solution is to allow the digitizer to come up

to operating temperature before using it.

o Repeatability :-

How close will be the first and second readout be?

Good digitizer should be repeatable to about 0.001 inch

o linearity :-

Digitizer is able to measure a specific distance of a correct value as the

puck moved over a large distance.

o Resolution :-

Digitizer should be able to record a increment of the space.

o Skew :-

Measure of the squareness of the results on a tablet

Do coordinates located at the four corners of your digitizer produce a

true rectangle, as intended?

Input of Raster Data, Vector Data or Both :-

No matter which approach is chosen for the GIS Input, but it is necessary

to determine at the outset whether you will be using a Raster or Vector

GIS.

Program, especially those that evolved primarily to handle remotely

sensed data, operate on the grid data structure where as others operate

primarily on vector data structure.

Line following scanners tends to produce Vector Output.

Drum scanners will produce raster Output.

Although conversation between Vector to Raster and Raster to Vector are

fairly common, there are something should be remember.



Page : 3

Conversation :-

o Vector To Raster :

Results are visually Satisfactory

But the attribute result is not satisfactory for each grid cell. This is

particularly true along the edge of area.

o Raster to Vector :

Result preserves the vast majority of attribute data.

Visual result will often reflect the blocky, step like format on the grid

cell from which the conversion proceeded.

Algorithms are available for smoothing this blocky appearance with

the use of mathematically based graphic technique called Splines.

That smooths out the jagged lines and sharp edges.

Reference Framework & Transformation :-

Digitizer input the map from aerial photography or other analog productof remote sensing equipment.

Maps are representation of a three-dimensional reference globe projectedon a flat surface.

There are three primary processes:

1. Translation.

2. Scale Change.

3. Rotation.

Translation :-

o Translation is simply movement of part or all of a graphic object to a

different location on the Cartesian surface.

o This is done by adding and subtracting the coordinate values necessary

for the X and Y coordinates for the object.

o In other words, the new X coordinate X’ for each graphic object will be

equal to the original X coordinate plus some value Tx, and the new Y

coordinate Y’ for each graphic object will be equal to the original Y

coordinate plus some value Ty.

X’ = X + Tx and Y’ = Y + Ty

Where Tx and Ty are the Translation amount.



Page : 4

Scale change :-

o Scale Change is also relatively useful because of the need to compare

differently scaled maps and to output in different scale as well

o This is done by multiplying the overall X coordinate extent by a scale

factor Sx, and each set of Y coordinates by Y scale factor Sy.

X’ = X * Sx and Y’ = Y * Sy

Where Sx and Sy represent the amount or percentage of scale change.



Page : 5

Rotation :-

o Rotation is used frequently during the process of projection and

inverse projection.

o It is also accomplish by using basic trigonometry.

o For new X coordinate locations, the new X coordinate X’ is found by

multiplying it by the cosine of the new angle (θ) and then adding that

value to the original Y coordinates multiply by the sin of theta (sinθ).

o The new Y coordinate location Y’ are found by multiplying the negative

of the original X value by the sin of the angle and again adding that to

the product of Y coordinate and sin θ.

X’ = X * (cosθ) + Y * (sinθ) and Y’ = -X * (sinθ) + Y * (sinθ)

Where θ is the angular displacement.

Map Preparation and Digitizing Process

We cannot begin the digitization process until we have provided the GIS

software with information regarding the projection, the grid and so on.

So it is important to developing the useful database.

In this process you should prepare the information beforehand and keep

it handy so that you will always know what it is and where to find it.

It is also a good idea, before you place your map on the digitizing table to

being the input process, to prepare a map by making appropriate marks

directly on the map document or on firmly attached clear plastic covering

to identify exact location you want to digitize.

If you know all the points, it is batter to digitize free hand.



Page : 6

Because digitization is a tedious work, you probably will digitize the map

in portions rather than all at once.

Registration Points / Tic Marks :-

Because you will probably digitizing the map in multiple sessions, and

sometimes you will have to remove the map to allow other t use the

equipment, you will need to tell the computer software where your

map area is and what its coordinates.

These first points, called Registration Points or Tic Marks, they will be

entered in digitizer inches as well as in map coordinates.

They also should be marked on the map as part of your map

presentation, so that you as well as your computer know your starting

points.

The registration points provide an outside frame for the

document.

Usually three points need to locate at the corners of a rectangle to

define the map area for the software. Some software can get by only two

points if they are located on diagonal.

In this case the software assume that the boundary is a rectangle and

infer the other two points.

The accurate location of the registration point is absolutely essential to

ensure good quality.

Special care should be taken to locate your Tic Marks precisely.

It is good idea to double-check these because if registration marks are

misplaced, virtually all the remaining digitization marks will be

erroneous.

And if you select that tic marks the software will provide you with an

error measurement using a method called Root Mean Square

(RMS) error.

The lower the RMS error the more accurate the digitization.

Other map preparations include a clear definition of the order in which

you intend to digitize; a systematic approach to identify the portion of

the map to digitize at each session has been mentioned.

It is also good idea to develop a method to identify which area have

already been digitize.

Most digitizing software provides editing capabilities to help you to

identify your mistake.



Page : 7

Most digitizing software includes a feature that allows for the

occasional shaky (7 �ƨT ^) hand, namely, a fudge factor sometimes called

fuzzy tolerance.

The inclusion of this feature reflects the assumption that you will not be

able to place the crosshair of digitize puck over exactly the same

location twice.

Small fuzzy tolerances are less forgiving of digitizing errors and many

results in gaps between points that were meant to be connected.

Final map preparation deal primarily with the tendency of the sourcematerial to shrink and swell with change in temperature and humidity(હવામાં [ pKȵ k5̆ \hR).The material you are about to digitize should be allow to

stay in the room for several hours, unrolled. It is good idea to avoid using unfold maps because creases reduce the

accuracy of such a document.

What to Input :-

We have some basic guidelines on how to digitize, especially how to avoid

errors during digitize, we can begin to select the appropriate data to

input.

A major factor guiding what cartographer put on the map and how it is

produce is the target audience, called Users.

The same can be said for producing the cartographic and geographic

database for map analysis within the GIS.

Different rules are there for deciding that what the inputs of any

particular GIS application are.

Rule : 1 (Purpose)

o To decide why you are building the GIS database in the first place.

o This will at least limit the input to coverage that is likely to be used.

Rule : 2 (Goal)

o Results to the first, is that you must define your goals as specially as

possible before selecting the layer.

o Because GIS begun uncertain guidelines to produce reasonable result

without substantial reworking and correctness.



Page : 8

Rule : 3 (Avoidance)

o Avoid the use of exotic sources of data when conventional sources

available.

o With known spatial information products, there will be multiple ways

to obtaining the available data in some cases.

Rule : 4 (Selecting the Best)

o If all your multiple data sources for the particular them are in

traditional form then Use the best, most accurate data necessary for

your task.

o If you and another team member are familiar with the given set of data

and can comfortably use it correctly and if it increases the utility of

accuracy of database, it should be applied.

Rule : 5 (Diminishing)

o Remember the law of diminishing return when deciding on data

accuracy levels.

o Difficult to separate categories over an area that is essentially all grain

fields.

o Computing the human resources needed to provide clarification would

increase the overall cost of system.

Rule : 6 (On Same Map)

o Whenever possible, and when the quality of data dictate it, you

should input these data as separate coverage of the same map sheet.

o Advantages :

Data are on the single map, you did not need to go to multiple map

sheets and then repeat all the preliminary steps in map

preparation.

Because the data are on the same map sheet they are already geo

referenced, reducing to perform this difficult task later.

Rule : 7 (Specification)

o Each them should be as specific as possible.

o The more specific a them, the easier it is to search if you need to know

something.

o When you perform an operation like overlay, it is easier to keep a track

of the processes if you are completely familiar with the data.



Page : 9

How Much to Input :-

“How much?” is the question that is related to the data types you will input.

For Example: As you prepare for your journey you will need to know how

much food to pack, not just what kind.

As with preparing packs for a trip, the input of data into a computer is a

sampling process.

In vector GIS, each line you input will likely have some curves. To produce

reasonable facsimile of the line using straight-line segments, you will

decide thousands of themes where to place the digitizer puck and where to

record the data.

The location of straight line can be recorded accurately with only two

points: one at beginning and one at ending.

Line and polygon complexity can be compared to information.

The more the line changes direction the more information it convey.

The higher the information content, the higher the sampling rate needs to

be known as Information Theory.

The idea of information content can also be applied to raster data. Once

again the general rule is this: the smeller the object to be identified in your

database, the smaller the grid cells need to be.

Resolution is the principle often determines the selection of grid cell size

for the entire database.

The Information Theory can also be applied to the input of raster data, but

keep in mind that grid cell are two-dimensional rather then one-

dimensional.

Whether in raster or vector, sampling is depending on he amount of area

covered by the map and the use for which the data are input.

Small scale maps, those covering large amount of space,

contain the much more abstract view of the land surface.

The amount of errors contain in the symbol is depending on the scale of the

map on which it placed.

Lines on the small-scale maps take up more land surface than

small-size line on the large-scale maps. This physical condition

called scale dependent error is an indication that the amount of error

is directly related to the scale of the map and need to be considered during

the map preparation phase prior to digitizing.



Page : 10

Methods of Vector Input :-

After preparing the map and putting it on the digitizing table, you will need

to use the digitizer puck to locate and record the registration marks.

Software documentation and the software itself will indicate the

requirements.

The software packages will indicate which numbered keys you need to

enter for specific object types.

Some numbered keys will be used to indicate the location of the point

entity, others for beginning and ending of line segment, and till other for

the closure of polygon.

Many digitizing errors, especially those made by invoices, are due to

pushing the wrong numbered button.

Methods of Raster Input :-

In the first place, we must decide how much area should be occupied by

each grid cell. This decision must be place prior to digitizing.

We must decide whether it is appropriate to use a method of encoding that

shortens the process, such as the run-length encoding.

Although compacting methods are good at reducing data, their use in

encoding may be even more important because of the reduction of the

input time. Once you have selected a method, you will need to decide how

each grid cell will represent the different themes that will occur.

Beyond the grid cell resolution, this may be the most important decision

you will have to make.

The entities and the attributes are entered simultaneously. This approach

required vector-to-raster conversation.

Most often difficulties extended from digitizing adjacent areas using

vector lines which are then converted to two separate polygons. In such a

case software must decide which polygon will contain the grid cell through

which the line runs.

The decision is sometimes based on the “Last Come, Last Coded” rule. This

is when the same line digitize first for one polygon, and then again for the

second polygon.

Different four methods are there :

1. Presence / Absence Method

2. Centroid-of-cell Method

3. Dominant Type Method

4. Percent Occurrence Method



Page : 11

1. Presence / Absence Method :-

A decision is made on the basis of

whether the selected entity exists

within the given grid cell or not-

hence the name is

Presence/Absence.

Advantages :

o This method is that decisions

are easy.

o No measurements are

necessary.

o Simple Boolean operator either “is” or “isn’t”.

o This method is useful for coding points and lines for grid systems

because these entities do not take up a large portion of a cell’s area.

o Thus if a road crosses through a grid cell, its presence is recorded

with an attribute code (a number); if does not, it is ignored.

2. Centroid-of-cell Method :-

Here the presence of an entity is

recorded only if a portion of it

directly at the center point of each

grid cell.

Clearly, this requires substantial

calculations, since each central

point will needed to be calculated

for each grid cell, and then the

object will have to be compared

with the location of that point.

You must evaluate each grid cell Centroid against each entity.

Therefore the use of this method will be restricted to polygon entity.

3. Dominant Type Method :-

This is the more common and best method of coding polygon data.

The presence of an entity if it occupies more than 50% of the grid cell.

Under the most circumstances, the decision is straightforward and

coding is reasonably representative of what is there.



Page : 12

It seems logical that if you are restricted to a single category for each

grid cell, the one that occupies the most space should be coded.

Computationally, this requires he

computer to determine the

maximum amount of each

polygon for each grid cell.

In two cases this approach is

satisfactory.

o First, there may be highly

irregular polygon shape. It is

difficult to decide such cases by visual inspection.

o Second, three or more polygons types coverage in an irregular

pattern within a single grid cell.

4. Percent Occurrence Method :-

This method is also used

exclusively for polygon data.

This idea is to give more detail,

not by coding just the existence

of each attribute but rather by

separating each attribute out as

a separate coverage and then

recording the percentage of the

area of each grid cell to

occupies.

For Example, a map of land use divided into urban and rural categories

would be separated into two more specific themes one urban and one

rural.

The percentage urban and rural would be recorded for each grid cell

This method offers the advantage of more detail data about each

attribute.

Remote Sensing Special Case of Raster Data Input :-

Remote Sensed data being either primary or secondary data sources.



Page : 13

Sources for Acquiring Remotely Sensed Data:-

The users may request specific satellite or aircraft over flights, for

predefined areas, at predefined time, with limited amount of cloud

cover.

Or user might have their own airborne or ground based sensing

system for acquiring their data. Under such circumstances, the

remotely sensed data would definitely be considered primary.

Some users currently employing some rather unique primary remote

sensing data approaches for small study areas, including the use of

tethered balloons with digital cameras.

These data are not dominant compare to the any other source of input,

such as traditional cartographic products, digital elevation data, digital

line graphs from topologic map, digital land used data and digital soil

data.

The raster appearance of remotely sensed data may give the impression

that GIS is software designed to manipulate these raster data.

Arial Photography :-

Aerial photography has long been a primary source of base map data

for many common products.

A special types of areal photograph deserves mention because the

images do not contain the scale, relief, and tilt distortion normally

characteristics of aerial photographs. These products called

orthophotographs are photographic image of the earth that

resemble maps in that they have a single scale.

The orthophotographs are subjected to process called Differential

Rectification, which involves point-by-point correction of the scale

and relief displacement normally caused by difference in elevation

between aircraft and the topography over which it files.

Technical Difficulties :-

Satellite Data require processing to remove geometric and

radiometric flaws resulting from the interaction of two moving

bodies. Techniques for correcting radiometric difficulties are readily

available with most digital image processing software, and necessary

equations are quite easily obtained.



Page : 14

For GIS input the major problem related to processing is a need to

obtain geometrically correct ground position for the imagery. The

geometric correction requires a number of Ground Control Points

(GCPs) within the image, to place it in a correct spherical coordinate

space on the surface of the earth.

Obtaining adequate GCPs may be quite difficult especially in area like

forest GPS field unit has no direct line of sight to satellite.

Care should be taken when using imagery lacking in GCPs because their

absence degrades the coordinate accuracy.

The issue of GCP accuracy should indicate the importance of coordinate

accuracy to the overall functioning of GIS.

Institutional Issues :-

1. General lack of availability of remotely sensed data.

Requires the user to be familiar with the process of obtaining the

data in the first place

Once the procedures are understood the problems could cover.

2. Difference in the atmospheric condition.

Change in the ground resulting from the flowering and dying off of

vegetation.

In other cases two or more contiguous satellite images may have to

be pieced together to cover a large study area completely.

3. The Hardware and Software Cost.

The ready availability of lower priced image processing software

running on standard computing hardware has improved this

situation considerably.

4. Data and in most cases the satellite systems themselves

were originally designed as experimental rather then

operational system.

5. Education and Training.

Today for both, GIS and Remote Sensing, the technical institutions

and image processing and GIS vendor companies offer programs

that provide extensive, hand-on experience with the use of particular

remote sensing or GIS system.



Page : 15

6. The final institutional problem of remotely sensed data

input features organizational infrastructure.

GPS Data Input :-

An increasingly important sources f primary data input to the GIS data

base is the GPS data.

It is often linked explicitly to the input of other primary data source such as

field data.

Software packages such as PENMAP allowed for the simultaneous input

of GPS locational information.

With the advent of the new handheld computer technology in the form of

PDAs with full color displays and software packages like ArcPad we are

now able to display digital map layers in full color in handheld unit, to link

those same layers with GPS coordinates, to include a complete set of field

notes.

The computer and GPS technologies have already been integrated to allow

out-of-the-box single-unit-solution.

Although such mobile solutions are available, they are not common place

because of the high price tag.

The obvious advantages of the mobile GPS/GIS solutions include :

1. Reduction in time and cost in development of geographic database.

2. Rapid GIS and remote sensing data verification.

3. Easy transfer of these data to full GIS database composes of both

primary and secondary data.

4. Incorporation of ancillary data and descriptions.

Secondary Data :-

The efficient method of building a GIS database is to limit the amount of

time and costs necessary to develop the database.

Fortunately, the digital databases are becoming increasingly available.

(Digital Elevation Model, Digital Line Graph)

The U.S. department of agriculture makes soil maps available in digital

form.

But the availability of database also introduces the other problems.

1. The physical format of media.

Countless hours can e spent trying to obtain digital data in the

proper format. So many formats of CD-ROM technology are

available.



Page : 16

You must be able to get the data in a format compatible with your

retrieval equipment.

Even if the media are same, data can be provided for numerous

formats.

2. Quality of the Data.

Though data vendor may provide easier access for data, service may

not provide data in original format.

Some data will be filled easy viewable errors, some systematic and

correctable, some not.

You need to be aware of quality control procedures used by each

vendor.

3. Use of an External Database.

External databases require information about their own content

such as: Data Dictionary. (Passive and Active)

Passive Data Dictionary might include scale, resolution, and the

name of data field s in the database, the codes used, and what they

mean.

Active Data Dictionary operates on the GIS database by

performing checks for correctly coded inquiry (Check Constraint).

For example, if your vector GIS database management system is set

up to allow only a four digits code for a particular entity, the active

data dictionary could check each inquiry to determine whether these

four-digit limitation is uniformly met. Such checks are quit useful for

preventing erroneous input.

Metadata and Metadata Standards :-

Although data dictionary are useful, they are not enough for the growing

set of GIS users, particularly where the data are shared among many users.

In such a case organizations are now adopting the needs for rigorous set of

standards that provide a detailed record of the dataset that they use and

share with others called Metadata.

Metadata describe in some detail the content, quality, condition, and other

important characteristics of the data thus helping others to locate and

understand.



Page : 17

The use of Metadata :-

1. Organize and maintain an organization's investment in the

data

In this case, the integrity of the data is maintained because they

contain and appropriate use of the data is cataloged.

2. Provide information to Data Catalog and Clearinghouse :

Among the most difficult tasks of GIS professionals is to identify

what data are available for their needs. The Federal Geographic

Data Committee (FGDC) is developing a network of geographic

data sharing called National Geospatial Data Clearinghouse

(NGDC) through which the metadata can be shared.

3. Provide information to aid in data transfer :

Any time the actual datasets are to be shared with others it is vital the

metadata be shared as well. This allows the user to process and

interpret data correctly and to match the data with its own data

holdings.

Following are the necessary things used to complete the

Metadata :

1. Identification

What is the database called?

Who developed it?

What geographic area are covered?

What category or them of data are included?

What if any restrictions are their for data access?

2. Data Quality

How good the data are?

How do the user know if the data are applicable for their use ?

How complete are the data?

Was data consistency verified?

3. Spatial Data Organization

What data model was use to encode spatial data?

How many spatial objects are there?

Are methods other then coordinates used to encode location?



Page : 18

4. Spatial References

Are coordinate locations encoded using Latitude/Longitude?

What, if any, projection is used?

What geographic datum was employed?

What parameters should be used to convert the data from one

coordinate system to other to preserve integrity?

5. Entity and Attribute Information

What specific information is included?

How is the information encoded?

Were codes used?

6. Distribution

Where and from whom do I obtain the data?

What formats are available?

What media types are available?

Can I obtain the data online?

How much does it cost?

7. Metadata Reference

When were the data completed?

Who complete the data?

chapter - 5 the input subsystem - ajay ardeshana · fundamentals of gis chap. 5 : the input...

Documents